Customer Repairs: Roberts Radios

 Typical of Roberts radios this is an R600 IF amplifier circuit board carrying old Mullard germanium pnp transistors. The transistors contain a compound used in their manufacture that eventually, after many years, causes a short circuit, of varying degree, between one or more internal parts and the metal case. Typically a leak may develop between the collector and the case of 200ohms. Usually the faulty transistor will work after a fashion but result in a drain across the battery leading to short battery lfe. Eventually the short gets worse, and may include the other electrodes, and the set no longer performs.

Often merely cutting the case lead, which connects that to ground will result in a 100% cure, but this may not last

 Getting to this circuit board to remove the transistors is extremely tricky as it involves detaching the push button assembly and cutting a large number of wires. Unless one is very careful marking these, or drawing a diagram, it is not easy to put everything back correctly.

Finally, if the transistors have been changed it is desirable to reset the various cores in the IF transformers. This is difficult as the circuits can readily burst into oscillation.

Roberts radios have always had a good reputation, at least with listeners. The sets look neat and are simple to operate. They have a chunky feel to them and having lots of wood in their construction gives them a pleasing tone unlike plastic sets whose tonal quality cannot compete. The phrase "good reputation" however surely cannot extend to anyone versed in their repair. Put aside the wooden case and what do we have inside? The circuit boards, loudspeaker and odds and ends that make up the bit that converts radio waves into sound. Admittedly, although the circuit design is first class and no doubt culminated after lots of esoteric mathematics, the way it's all put together is absolutely awful. Most Roberts radios have a nice cabinet and excellent circuit design, but rely on a rat's nest of wiring to hold everything together.

Try to change a transistor on a circuit board carried in a metal box that's buried in a mass of wiring and you'll see what I mean. Investigate a few things and you'll soon find a couple of wires have become detached. Both will have the same colour sleeving and each has a fragile single wire core and very little by way of clues as to where they should go back. Fortunately such is the pleasure a good Roberts radio can give it all seems to be worth the trouble.

R600, S/No 77437, Repair No. 3

 It's amazing this set had lasted so long! Someone else would have chucked it out... however the owner is a devout Roberts devotee.

The on/off switch wasn't working and the small contacts, used for speaker muting, on the headphone jack were open circuit. In addition the dial cord was broken and wrapped around the mechanism and the tuner unit was badly bent out of position. This latter problem seemed to be the reason for the dial cord coming adrift. I also found a break in the printed circuit which seemed to date back to manufacturing. This had reduced the overall sensitivity of the radio (from new!). Two audio driver transistors were very leaky and below par. The battery was well on the way to getting flat but still works on low volume. The ferite rod was detached from its mounting and the VHF aerial was broken off. The volume control knob was broken and a chassis fixing screw was missing.

After correcting all the faults I fitted an external mains power supply using a 13Amp plug unit putting low voltage AC into the radio via a phono socket mounted in the perforated plastic rear panel behind which I fitted a rectifier and regulator with a battery/mains diode switch. The set worked very well for 6 years....

R700, Repair No. 13/14

 On hearing of my mains conversion of her friends Roberts 600 I was asked to modify a Model 700. An overhaul was first required though and I found; a very gummy wavechange switch; a shorted 10.7MHz IF transistor; a shorted 470kHz IF transistor; broken earth return track; unsoldered capacitor (C27) missed in manufacturing!; lead to C31 cut and the capacitor faulty...very strange..and a note stuck to the side of the metal RF box with the legend "noisy" ....even stranger...?? This must have been rejected at Roberts then picked up inadvertently and used in the production line.

I replaced an AF114 and an AF117, dismantled the metal box and bent a diode (MR7) away from the inside wall; soldered C27; fitted a new 100uF for C31; cleaned the wavechange switch assembly and finally glued together one of the side panels which was split.

The battery eliminator was next. The 18 volts supplied by two batteries in series was a problem which I overcame by using a standard external 13Amp plug unit supplying 9 volts of AC to a voltage doubler inside the radio (2 capacitors, 2 diodes and a couple of surge limiting resistors). Spare contacts on the on/off switch were used to switch the incoming AC to the doubler and a pair of germanium diodes complete the modification, switching from mains to battery when AC power is lost. Until recently, here in the New Forest Area, we would lose our mains supply at the drop of a hat. Bare wires were routed through trees (which grow constantly and also move about in wind). Not long ago the wires were all replaced with insulated cables and much improvement followed until very recently when we seem to be again plagued by cuts. The man at the head office seemed to think it was swans flying into the wires! This must be a new game or he couldn't think of anything more sensible!!

R505, Repair No. 42

 A flat battery didn't help to divert blame from the radio because most of the Roberts I've had are riddled with faults. This was no different. Bare wires in the vicinity of the tuner box were intermittently shorting against each other, poor soldering abounded and the speaker cone coil was rubbing on the inside of the magnet slot. After resoldering connections, rerouting bare wires and fitting a new speaker and a new battery, results were wonderful. The new 3 watt speaker is infinitely more robust than the original one which looks like it was designed for 3 microwatts. A request for modification to mains was easier than usual as this model already carries a socket for the purpose.

RT1, Repair No. 691

 This 1958 model arrived with a request for a mains power supply and was unusual as it actually worked from its battery. The power supply followed the same lines as Model 600 except this one requires 6 volts rather than 9. All that was required apart from the battery eliminator was realignment of the IF coils after which sensitivity was excellent.

R505, S/No 96661, Repair No. 874

 Distorted sound wasn't a duff loudspeaker. I connected a scope and a signal generator and found even harmonics being produced in profundity at the speaker terminals. This turned out to be a dry joint at the connectioons to one of the two push-pull output transistors. A search revealed lots of other poor soldering. Compared with far east offerings this make of radio has an abyssmal standard of mechanical design and construction. Electronic design however is first class and performance of a nice new model that passed QA must have been very good. Over the years poor build quality and the old Mullard AF117 transistors let it down through hopeless reliability.

RM20, S/No 28996 Repair No. B188

 The complaint was noisy operation and loss of sensitivity. The former was cured by applying switch cleaner to the volume control and the wavechange switches and the latter by realignment of the 470kHz IF, the RF coils and the position of the coils on the ferrite rod.

R500 S/No 20593, Repair No. C556

 No sound because the first RF transistor had an internal short circuit. Cutting the connecting wire to the case cured the problem. Two dust cores were damaged and after replacing these, realigning the IF to 470Khz and realigning the front end all was well.

R309, Repair No. C599

 The display was out because of a faulty microprocessor. Sadly the estmated cost of repair exceeded the value of the radio.

RC45, Repair No. D039

 This radio/cassette had a faulty integrated circuit. Fortunately the cost of a replacement chip was low and its fitting was almost straightforward.

The dismantling was just as long winded as older Roberts models; dismantle radio, detach dial assembly, detach cassette assembly and finally detach radio circuit board. I was then able to remove the LA4160 equaliser-power amplifier chip and fit an equivalent KA2213.

RT1, S/No.T7793, Repair No. D472

 All that was wrong was a broken wire at the battery connector. On test though I found sensitivity was poor. This was sorted out after the front end and the IF amplifier had been realigned.

R600, S/No.45691, Repair No. E094

 This set wasn't receiving signals because the old "AF" coded transistors had short-circuited. I removed the metal IF amplifier box and extracted the circuit board. After removing the transistors, in each example, I cut off the lead connecting to the case. The devices tested OK so I resoldered them into place and refitted the tuner. Battery consumption measured less than 20mA and the set was very lively without the need for realignment.

RT24, S/No.19678, Repair No. E100

 This set was finished in a very attractive shade of blue and was envied by all that saw it in the workshop. I understand it had been presented as a long service award to the mother-in-law of a customer and had to be fixed.

Fortunately the fault was quite simple to diagnose. The set only came on for a second or two as the on/off button was being pressed because the switch contacts had worn. In this model the switch is in a gang carrying the wavechange switches so it is not that easy to replace. A couple of solutions were considered but the most straightforward was to use an unused section on the existing on/off switch. The switch is a two pole changeover affair but only one pole is used. As one of the unused poles contacts had been used as an anchoring point for ferrite rod wiring, the connections to this were moved and the new pole paralleled up with the old. Easy but needed thinking about before diving in and removing the whole assembly etc. I noted that if the contacts again failed the switch could be wired back-to-front as an "off/on" switch as both poles weer working in the changeover position.

R404, S/No.19886, Repair No. E125

 Again the stock fault of duff AF117 transistors. After removing the metal IF box and testing the transistors I cut their case leads and refitted two of three that tested OK. As the IF stages are prone to oscillation if a transistor with a different spec is fitted I decided to pinch the first RF transistor which is also an AF117. I removed this, cut its case lead and fitted it in the IF box. I used a small germanium transistor with a reasonable HF spec as a replacement for the RF one. Unfortunately after reassembly the set was really deaf. I found, by injecting an IF signal at quite a high level into the aerial socket that someone had been twiddling the small slugs in the IF transformers and some of these were broken. After taking out the IF box and removing its circuit board and then the small transformers I was able to extract the cores. It's a pain as there isn't a hole through the circuit board and the slugs have to be pushed out rather than unscrewed out. This problem is that thin rubber is used to provide friction for the cores which can be screwed up or down only if the rubber hasn't perished. I used strips of thin card in place of the rubber and used the damaged cores upside down to re-use them. Once everthing had been reassembled it was quite difficult finding the right settings as all the cores had been moved from the correct tuning positions.

Two electrolytic capacitors in the audio amplifier had been damaged by the phantom twiddler as well. Once these had been replaced the set worked reasonably well. Although the set uses two 9 volt batteries in series it works tolerably well with just 9 volts.

RC15, Repair No.E165

 This radio needed only a squirt of switch cleaner on its volume and tone controls. This model was made in Taiwan and looks just like any other far east import except removing its back was a tussle because of weak mechanical design. When I finally got it off I found what looked like a steel rivet inside lying loose in the works. I put it on one side thinking it must have been rattling around a bit. Getting the back off reminded me of removing a silencer from a Saab 96. When I'd been at it for hours I rang the local dealer who said you just had to jiggle it till it fell off! You also had to have the car at 45 degrees to the horizontal of course. Getting a new one on was even harder than getting the old one off, just like the back of this Roberts. When it was done one side of the handle fell off. That rivet thing was supposed to be holding it on. Of course the back had to come off to get the rivet back in!!

R200, Job No.H055

 The owner of this set wished to operate it from the mains. To circumvent difficulties with safety the best way to provide mains operation is from one of those "13 Amp Plug" type of units. Modern units provide a stabilized 9 volt output, free from hum and are fine for medium and long wave sets.

The R200 carries a car aerial socket on its hinged rear panel and the easiest way to keep the set's originality is to remove this socket and fit a 2.5mm circular socket (CPC stock number CN06488) after drilling the hole to 12.5mm. A typical power supply (CPC stock number PW00503) provides a regulated 9volts at 300mA and this should be modified by cutting off its universal connector and soldering a 90 degree cranked 2.5mm plug (CPC stock number CN04251) to its cable. I always either cut off the existing plug and fit a new one as the universal types can be readily reversed, and unless extra safety components are fitted, will wreck the radio. The adage "if it can happen.. it will happen" is a useful guide.

If the set had an FM band there may have been a problem with inter-modulation. This surfaces as a loud hum tuning with broadcast stations and is sometimes tricky to overcome. Decoupling the power lead with RF capacitors can help but occasionally I have been forced to fit a low voltage AC power unit, with rectification and smoothing inside the set, so that adequate decoupling can be provided.

The power socket I used in the R200 has an integral switch so that when the power unit is plugged into the radio the internal battery is disconnected. In the past I've not always used this technique, for example when a customer wished her radio to carry on working after a power cut (which used to be very frequent in the New Forest), I used a pair of germanium diodes arranged to allow whichever supply was highest to take priority. The mains voltage was of course set slightly higher than the terminal voltage of a new 9 volt battery. If there was a power cut the battery diode operated in an instant and no loss of program occurred. The use of Ge diodes minimised voltage drop to a minimal level. Nowadays Schottky silicon diodes may be used as these provide a very low "on" forward resistance.


Return to customer repair page